SIENTIAPDE-1646

Refactor ModelMetrics to utilize DriftAnalysis for drift detection

- Replaced ModelAnalysis with DriftAnalysis in the ModelMetrics class to enhance drift detection capabilities.
- Updated method signatures and documentation to reflect the changes in target_name and return values.
- Adjusted data handling to ensure compatibility with the new analysis methods and improved clarity in the drift metrics dataframe preparation.

e2e/scenarios.md

@@ -232,81 +232,87 @@ Source: `e2e/test_minio_offload.py`
 ## 5. Drift Workflow Scenarios
 Source: `e2e/test_drift.py`
 
-The drift suite mocks `sientia.ModelAnalysis.ModelAnalysis` (not installed; see
-`CODE_ISSUES.md` issue #1) through the controllable `_FakeModelAnalysis` stub
-exposed by the `model_analysis_stub` fixture. The `mlflow_repository_stub`
-provides the reference-data CSV via `download_artifacts`. Every scenario asserts
-postgres rows in `predictions_schema.drift` against this canonical schema:
+The drift suite drives the **real** `sientia_model.analytics.drift_analysis.DriftAnalysis`
+analyzer (no stubs / mocks). Each scenario exercises the full pipeline:
 
-`id, model_id, feature, method, value, drift, chunk, timestamp, timestamp_end, accurate, created_at, updated_at`.
+```
+laborious_data (Postgres) -> load_custom_query
+                          -> calculate_drift (DriftAnalysis univariate + multivariate)
+                          -> export_data_to_postgres (sientia_data.drift_metrics)
+```
+
+The `mlflow_repository_stub` provides the reference-data CSV via
+`download_artifacts`, and tests assert postgres rows in
+`sientia_data.drift_metrics` against this canonical schema:
+
+`id, model_id, feature, method, value, alert, chunk_index, chunk_start_date, chunk_end_date, accurate, timestamp, created_at`.
+
+Tests assert behavioral / structural properties (column presence, NOT NULL
+constraints, business-key invariants like uniform `timestamp` and stamped
+`model_id`) rather than exact numeric drift scores, since those depend on
+the real analyzer implementation and the synthetic data fed in.
 
 ### 5.1 Happy paths
 
 #### D.1.1 Full pipeline persists all columns with reference data
-**Summary**: ModelAnalysis returns a deterministic drift dataframe; the
-reference CSV is downloaded from the MLflow stub.
+**Summary**: 10 minutes of target data are inserted; a 10-row reference CSV
+is configured via the MLflow stub. The `DriftAnalysis` runs end-to-end.
 
 **Expected Outcome**:
-- One row per `(chunk, feature, method)` plus a `multivariate` block per chunk.
-- Every drift column is populated and `accurate=True`.
-- `timestamp_end` preserves the high-precision string (`HH:MM:59.999999999`).
+- One row per `(chunk_index, feature, method)` plus a `multivariate` block
+  per chunk is persisted.
+- Every column in the DDL is populated; `feature` is the only nullable column
+  per the new schema.
+- `accurate=True` for every row (reference path).
+- All three default univariate methods reach the analyzer.
+- `model_id` is stamped as `text` and uniform across rows.
+- `timestamp` equals `max(target_data.timestamp)` and is uniform across rows.
+- `chunk_start_date` / `chunk_end_date` are persisted as ISO text and ordered.
 - `p_value` is dropped before persistence.
-- `drift` flags propagate per `(feature, method)` configuration.
 
 #### D.1.2 30% fallback when reference data is unavailable
-**Summary**: `get_reference_data` fails alias resolution and returns `None`;
-`calculate_drift` uses the first 30% of target rows as reference.
+**Summary**: MLflow alias resolution is forced to fail so
+`get_reference_data` returns `None`; `calculate_drift` falls back to the
+first 30% of target rows as reference.
 
 **Expected Outcome**:
-- Persisted rows carry `accurate=False`.
+- All persisted rows carry `accurate=False`.
 - A `MODEL_METRICS_REFERENCE_DATA_WARNING` notification is emitted to MongoDB.
 
-### 5.2 Filtering / dedup invariants
-
-#### D.2.1 Deduplication and `p_value` removal
-**Summary**: ModelAnalysis returns duplicate `(timestamp, method, feature)` rows
-plus a `p_value` column.
-
-**Expected Outcome**:
-- Duplicates are collapsed keeping the first occurrence.
-- `p_value` is absent from the persisted rows.
-
-#### D.2.2 Out-of-range timestamps filtered
-**Summary**: Drift rows whose timestamps are not present in the target window
-must be discarded before persistence.
-
-### 5.3 Failure paths
+### 5.2 Failure paths
 
 #### D.3.1 Empty target data short-circuits the workflow
-**Summary**: `load_custom_query` returns no rows; ModelAnalysis is never
-instantiated and no drift rows are written.
+**Summary**: `load_custom_query` returns no rows.
 
-#### D.3.2 ModelAnalysis raises during dataframe assembly
-**Summary**: `get_drift_metrics_dataframe` raises. The activity catches the
-error, sends a `MODEL_METRICS_GET_DRIFT_METRICS_ERROR` notification, and the
-workflow completes without persisting drift rows.
+**Expected Outcome**:
+- The workflow returns early and writes nothing to `sientia_data.drift_metrics`.
 
-### 5.4 Configuration paths
-
-#### D.4.1 Default drift metrics propagated to analyzer
-**Summary**: Omitting `drift_metrics` defaults to
-`['kolmogorov_smirnov', 'jensen_shannon', 'wasserstein']` and forwards the
-exact list to `detect_univariate_drift`.
+### 5.3 Configuration paths
 
 #### D.4.2 Invalid `chunk_period` raises ValueError
 **Summary**: Anything other than `min` / `s` is rejected by `calculate_drift`.
 
-#### D.4.3 `chunk_period='s'` keeps seconds in timestamp filtering
-**Summary**: Truncated `YYYY-MM-DD HH:MM` rows are filtered out when chunking
-runs at second granularity.
+**Expected Outcome**:
+- The workflow surfaces the `ValueError` ("Invalid chunk period: ...").
+- No rows are persisted.
+
+#### D.4.3 `chunk_period='s'` preserves seconds in `chunk_start_date`
+**Summary**: Target data spans two minutes with samples at second-30
+boundaries; the activity is configured with `chunk_period='s'`.
+
+**Expected Outcome**:
+- At least one persisted `chunk_start_date` carries `seconds=30`, proving
+  that the analyzer chunked at sub-minute granularity and the ISO-text
+  serialization preserved the boundary.
 
 ---
 
 ## 6. Simple Metrics Workflow Scenarios
 Source: `e2e/test_simple_metrics.py`
 
-Validates `predictions_schema.simple_metrics_data` columns:
+Validates `sientia_data.simple_metrics` columns:
 `id, model_id, metric, value, timestamp, data_size, interval_minutes, created_at`.
+Note: ``timestamp`` is now nullable per the new DDL and ``model_id`` is ``text``.
 
 ### 6.1 Happy paths
 
@@ -339,8 +345,10 @@ the workflow exits before `calculate_simple_metrics` and writes nothing.
 Source: `e2e/test_minimal_retrain.py`
 
 The MLflow registry is fully mocked (no real artifacts in test container).
-Validates `predictions_schema.retrain_reports` columns:
-`id, model_id, model_name, timestamp, status, version, mlflow_run_id, mlflow_experiment_id, created_at`.
+Validates `sientia_data.log_retrain` columns:
+`mlflow_experiment_id, mlflow_run_id, model_id, model_name, status, timestamp, version`.
+Note: the new DDL drops the legacy ``id`` and ``created_at`` columns,
+``mlflow_experiment_id`` is now ``int8`` and ``model_id`` is ``text``.
 
 ### 7.1 Happy path
 
@@ -350,7 +358,7 @@ the new version is promoted to the `production` alias.
 
 **Expected Outcome**:
 - Report row has success status, `version='7'`, `mlflow_run_id='retrain-run-id'`,
-  `mlflow_experiment_id='experiment-id'`.
+  `mlflow_experiment_id=4242` (`int8`).
 - `mlflow.log_artifact` is called with the input CSV.
 - `promote_to_alias` is called once with the resolved version and alias.